Every leftover from the last orbital sentence has a different object. Hostess is a bag on a shelf. Two Falcon 9 rideshares are the booked hardware. Fifty Mini Laser terminals are a purchase order. BlueBird 7 is a $30 million line. Otter is a scoop that wants to live in the air that is supposed to bring the bird down. Flatten those into one space trade and you get a false headline. Do not flatten them.
DARPA Otter’s VLEO band. Redwire $44m Phase 2, 19 Nov 2025, SabreSat delivery. Voyager AMS, 21 May 2026. Kreios + NanoAvionics MP42, 4 Aug 2026, 150–300 km.
Three altitudes, one atmosphere
The dates are already on paper. 19 November 2025 on a $44 million Phase 2. 21 May on a Voyager package. 4 August on a bus selection. A date is not a flight.
Otter is the exposed scoop
The object the catalog actually produced is a government scoop and a commercial bus selection, not a constellation that already lives at 90 kilometres. DARPA's Otter program will fly an air-breathing electric thruster in very low Earth orbit, 90 to 450 kilometres, and stay there by scooping the air. Redwire, 19 November 2025: a $44 million Phase 2 to finish the SabreSat orbital drone, manufacture it, and deliver it to launch. Voyager, 21 May 2026: the acceleration package for the same bird. Electric Propulsion Laboratory took a $6.7 million 2024 award for the intake and thruster. Voyager's AMS has flown on NASA's Magnetospheric Multiscale. DARPA has not published a flight date. Phase 2 money is for manufacturing and delivery. A bird that has not been delivered has not been flown.
Kreios Space and Kongsberg NanoAvionics, 4 August 2026: the first commercial air-breathing electric propulsion satellite, a roughly 200 kilogram MP42, 150 to 300 kilometres, a yet-to-be-named optical payload, sub-metre VNIR. Starcloud is arguing about 2028 versus 2029 at 500-plus kilometres. These people are arguing about whether you can live in the atmosphere.
A satellite above 500 kilometres is a compromise with drag. Imaging resolution, latency, and link budget all get worse as you go up. VLEO is the opposite compromise: you get closer, you get sharper, you get a better phone link, and you fall. Propellant is finite. Otter's sentence, on DARPA's own page, is that ambient air is a virtually unlimited supply of propellant if you can ionise it and throw it out the back. The program manager is Michael Mulqueen. The end state is a flight longer than one year that DARPA is calling an orbiting wind tunnel, because you cannot recreate free-molecular VLEO on a bench.
It is a shoreline story. The first customers DARPA and Redwire name are defence, intelligence, and communications. Tom Campbell, Redwire space-missions president, on the $44 million award: higher-performance missions at lower altitudes, better perception, more revisit, less latency. Kreios says the same nouns in Spanish and Lithuanian: sharper images, better link budgets, 150 to 300 kilometres. If you can stay there, you do not need an 88,000-satellite compute file to change a picture. You need a scoop and a thruster that does not choke on atomic oxygen.
Neither object is in Starcloud's 88,000-satellite file. Neither object is in SpaceX's million-sat file. Both objects, if they work, consume more fairings, not fewer: VLEO life without a working scoop is 30 to 90 days. A refill every season is a rideshare book. A scoop that runs for a year is a different book.
Otter is a U.S. government object. Redwire is the prime. Phase 2 money is for manufacturing and delivery, not for a press-kit rendering. Kreios's 4 August date is a bus selection, not a launch. NanoAvionics MP42, about 200 kilograms wet, optical payload still unnamed. First-ever is Kreios's word. It will be true the day the thruster breathes on orbit and the altitude holds. It will be a sentence if the bird circularises and the intake starves.
Do not promote a scoop into a constellation. Do not ignore the scoop because the award is $44 million and the bus is still on the ground.
What a scoop cannot do
Cross-asset does not care which intake sits on a bird. AI abundance cares only if the middle's hosting cost becomes an altitude cost. It has not, on the invoices those pages already named. The orbital-commons page already named Otter and Kreios. The two-sats page already named the 500-plus kilometre argument. This piece is the scoop.
A refutation would look like this: Otter on orbit for more than a year, with a public note that the intake produced usable thrust, by the end of December 2028, means the shoreline held. A Kreios, or named peer, air-breathing bird that holds 150 to 300 kilometres for 30 days without a chemical dump, by the same date, means the commercial bus held. A year of Phase 2 slides and a bus that never leaves the hall means the award stayed an award. What this does not say: that LEO compute is fake, or that VLEO replaces Starlink. It says the unused orbit is the one everyone filed above. The shoreline is 90 kilometres. The propellant is the air. There is no recommendation to own a scoop.
The honest uncertainty is which object gets a number first. If Otter flies for a year and the intake produces usable thrust, the scoop has a flight. If Kreios holds 150 to 300 kilometres for 30 days and Otter is still a Phase 2, the commercial bus moved and the government scoop did not. If a year of slides pass and neither bird leaves the hall, the file moved and the shoreline did not. Those three sentences can live in the same year. Forty-four million is a fact about one award. 4 August is a fact about one bus. Neither is a year in the air.
Satellite bus on the cleanroom floor.
Wafer and a radiator plate.